Fault recovery method and apparatus

A a multi-processor system including a main storage and a plurality of virtual machines which are assigned to a plurality of processors under control of a host operating system and a plurality of guest operating systems and which operate on the processors associated therewith in which a fault occurring in one of the processors is detected to recover functions of the system, comprising:

• a table to set therein recovery attributes for failure to each of the virtual machines;
• circuit means for storing, when one of the plural virtual machines assigned to the plural processors under control of the host operating system and plural guest operating system detects during its operation an event of occurrence of a fault in one of the processors, data and status information of virtual machines assigned to said one of the processors and status information of said one of the processors from the processor respectively in a virtual machine save area and a real machine save area of the main storage;
• means for reporting the event of occurrence of the fault to at least one of the plural processors;
• means for storing fault information of said one of the processors in the main storage by a fault recovery circuit of one of the processors having received the fault report and causing an interruption in the plural processors other than the processor in which the failure occurred;
• means for restarting by one of the interrupted processors an interrupted process of the processor according to the status information of the virtual machines and the status information and fault information of said one of the processors in which the fault occurred obtained from the main storage and causing an interruption in the host operating system;
• and
• means for retrieving by the host operating system having received the interruption one of the virtual machines assigned to said one of the processors according to the processor status information in the real machine save area and assigning the plural virtual machines to the plural processors according to the recovery attributes for failure set to the retrieved virtual machine.

A fault recovery method, for use in a multiprocessor system including a main storage and a plurality of virtual machines which are assigned to a plurality of processors under control of a host operating system and a plurality of guest operating systems and which operate on the processors associated therewith, of detecting a fault occurring in one of the processors and thereby recovering functions of the system, the method comprising the steps of:

• setting recovery attributes for failure to each of the virtual machines;
• operating at least one of the plural virtual machines assigned to a plurality of processors under control of the host operating system and plural guest operating systems;
• storing, at detection of occurrence of a fault in one of the processors, data and status information of virtual machines assigned to said one of the processors and status information of said one of the processors from the processor respectively in a virtual machine save area and a real machine save area of the main storage;
• reporting the occurrence of the failure to at least one of the plural processors;
• storing fault information of said one of the processors in which the fault occurred in the main storage by a fault recovery circuit of the processor having received the report of the fault and causing an interruption in the plural processors other than said one of the processors in which the fault occurred;
• restarting, by the processor having received the interruption, an interrupted process of the processor according to the status information of the virtual machines and the status information and fault information of said one of the processors in which the fault occurred and causing an interruption in the host operating system, the information being obtained from the main storage;
• and
• retrieving, by the host operating system having received the interruption, one of the virtual machines assigned to said one of the processors in which the fault occurred according to the processor status information in the real machine save area and then assigning the plural virtual machines to the plural processors according to the recovery attributes for failure set to the retrieved virtual machine.

A fault recovery method for use in a multiprocessor system including a main storage, a plurality of logical partition resources which are assigned to a plurality of processors under control of a host operating system and a plurality of guest operating systems and which operate on the processors associated therewith when operation modes respectively thereof are altered from a host mode to a guest mode, and a plurality of virtual machines having a multi-processor configuration in each of the logical partition resources, the method detecting a fault occurring in one of the processors and recovering functions of the system, the method comprising the steps of:

• setting recovery attributes for failure to each of the virtual machines;
• operating at least one of the plural virtual machines in the logical partition resources assigned to a plurality of processors under control of the host operating system and plural guest operating systems;
• storing, at detection of occurrence of a fault in one of the processors, data and status information of virtual machines assigned to said one of the processors and status information of said one of the processors from the processor respectively in a logical partition resource save area and a real machine save area of the main storage;
• reporting the occurrence of the failure to at least one of the plural processors;
• storing fault information of said one of the processors in which the fault occurred in the logical partition resource save area of the main storage by a fault recovery circuit of the processor having received the report of the fault and causing an interruption in the plural processors other than the processor in which the failure occurred;
• restarting, by the processor having received the interruption, an interrupted process of said one of the processors in which the failure occurred according to the status information of the virtual machines and the status information and fault information of said one of the processors in which the failure occurred and causing an interruption in the host operating system, the information being obtained from the main storage;
• and
• retrieving, by the host operating system having received the interruption, one of the virtual machines assigned to said one of the processors in which the failure occurred according to the processor status information in the real machine save area, then arranging the attributes, and thereby setting the virtual machine to an operation halt state.